Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

First physiological test for autism proves high accuracy in second trial

news.rpi.edu

Post reply on HN
Print view View original

41–50 of 110 posts

Re: First physiological test for autism proves high accuracy in second trial

#41

Post by vorotato »

jmcgough wrote: post #5
scythe wrote: post #4

Very surprising result. I don't think I've ever heard of a blood test for a mental disorder -- at least, not one that actually works.

Be careful about that kind of wording. I and friends who're on the spectrum don't consider it a "mental disorder". It's simply a physiological difference, one that theoretically can be tested for.

If it's not affecting you in a negative way then it's not a disorder. If it is then it can be. If it doesn't have serious negative impacts then you're just non-neurotypical which is fine. People have the same opinions around ADHD but when you need accommodation or treatment it's worth classifying it as something that needs some level of accommodation or treatment.

Re: First physiological test for autism proves high accuracy in second trial

#42

Post by TheSpiceIsLife »

hhh wrote: post #38
TheSpiceIsLife wrote: post #35

Earlier quoted context omitted.

> spectrum is of infinite width and infinite height If this were true, everyone is diagnosable.

It was more for the aesthetic of the sentence, but if you have a spectrum including 'normalcy' that is infinite width and infinite height, with x being ability (a combined measurement of communication ability, sensitivities, etc) and y being level of function, it is possible to include 'normalcy' as it is defined at the farthest top right corner. This could even have more dimensions to it, represented as 0-1 on each.…

That seams reasonable enough.

I have two thoughts: a) this risks become a personality test, which are notoriously poor as a medical diagnosis tool, and b) I don’t think we (society) should be in a hurry to slap a medical diagnosis on everything.

Re: First physiological test for autism proves high accuracy in second trial

#43

Post by vorotato »

robrenaud wrote: post #37

Statistical reporting is often pretty frustrating. > “We are able to predict with 88 percent accuracy whether children have autism,” said Juergen Hahn, lead author > It is estimated that approximately 1.7 percent of all children are diagnosed with ASD def has_autism(child): return False apparently has 98% accuracy.

I would hope that there's more rigor than that. I mean if you're willing to be that dishonest with your sample selection you could select only people who don't have autism and get 100% accuracy.

Re: First physiological test for autism proves high accuracy in second trial

#44

Post by icegreentea2 »

robrenaud wrote: post #37

Statistical reporting is often pretty frustrating. > “We are able to predict with 88 percent accuracy whether children have autism,” said Juergen Hahn, lead author > It is estimated that approximately 1.7 percent of all children are diagnosed with ASD def has_autism(child): return False apparently has 98% accuracy.

It's 88% correct classification on a validation data set consisting of 154 patients diagnosed with ASD. Their validation dataset did not contain a "typically developing" population since there's only one dataset for that data and it was consumed in their model generation. Their training data resulted in a 5% misclassification rate of their controls - that is 5% of patients that were not diagnosed with ASD in current tests would be classified as having ASD in this scheme.

The full paper is here: https://onlinelibrary.wiley.com/doi/10.1002/btm2.10095

Re: First physiological test for autism proves high accuracy in second trial

#45

Post by lujim »

BadCookie wrote: post #16
JumpCrisscross wrote: post #9

Earlier quoted context omitted.

> Be careful about that kind of wording I have friends on the spectrum. I happen to agree that mildly autistic behaviour is more akin to a personality difference than a disorder. That said, I don't think it's fair to admonish someone for using the colloquial norm. The comment might read better if stated less as "careful, you're doing something wrong" and more in the tone of "here is another viewpoint you may find hel…

To add to your comment, I wish that high-functioning ASD people would stop ignoring the more severe end of the spectrum. If you are high-functioning, then maybe ASD just makes you quirky and different. But for the lower functioning people, ASD can profoundly affect their lives in extremely negative ways. For example, my son almost starved himself to death because of sensory difficulties. He is four years old and stil…

> I've had people get mad at me for saying that I have a son with autism when I apparently should have said that I have an autistic son.

It has become exhausting trying walk the ever shifting minefield of acceptable verbiage. I'm not sure how you would apply this to other situations. If you have a son with red hair should he be called your red headed son? That sounds more offensive right?

Re: First physiological test for autism proves high accuracy in second trial

#46

Post by lujim »

gowld wrote: post #34
throwawaymath wrote: post #29

Earlier quoted context omitted.

No, "disorder" is explicitly clinical. The words "retarded", "dumb", and "stupid" are pejorative. Calling autism a disorder is an accurate paychiatric description, not a value judgement designed to insult a person's identity. An insult can be targeted at a person's autism, but not simply by describing it according to clinical terminology. I sympathize with people who don't believe autism should be considered a disord…

"retarded" and "dumb" are also clinical terms. They only stopped being clinical terms because people attached connotations. https://en.wiktionary.org/wiki/euphemism_treadmill

Yes this happens over time as clinical terms are used in a derogatory manner over a period of several decades. This will always happen and some clinical terms will continually need to be revised over time. It doesn't mean that "disorder" isn't appropriate in 2018. It just means it may not be appropriate in 2058.

Re: First physiological test for autism proves high accuracy in second trial

#47

Post by hhh »

TheSpiceIsLife wrote: post #42
hhh wrote: post #38

Earlier quoted context omitted.

It was more for the aesthetic of the sentence, but if you have a spectrum including 'normalcy' that is infinite width and infinite height, with x being ability (a combined measurement of communication ability, sensitivities, etc) and y being level of function, it is possible to include 'normalcy' as it is defined at the farthest top right corner. This could even have more dimensions to it, represented as 0-1 on each.…

That seams reasonable enough. I have two thoughts: a) this risks become a personality test , which are notoriously poor as a medical diagnosis tool, and b) I don’t think we (society) should be in a hurry to slap a medical diagnosis on everything.

I don't agree with slapping a diagnosis on anything either, but I think it could be a reasonable tool for people who are in the process of being diagnosed, or looking to better understand their character traits combined while going through the sometimes complicated process of being diagnosed.

Re: First physiological test for autism proves high accuracy in second trial

#48

Post by lujim »

smt88 wrote: post #22
cosmojg wrote: post #6

Earlier quoted context omitted.

Doesn't that reasoning apply to any disorder? Medically, a disorder is simply "a disruption of normal physical or mental functions." So, having any disorder makes one abnormal, unique, special, and possibly above average or below average, but certainly not average. "Disorder" and "special physiological difference" are synonyms.

Is that the psychiatric definition of disorder? My understanding is that it has to be disagreeable to the person who has it or put others at risk (in extreme instances). So for example, someone with synesthesia might be considered disordered only if it prevents them from being happy or taking care of themselves, while many people with synesthesia seem to enjoy it.

A clinically important syndrome; that is, it's a collection of symptoms (these can be behavioral or psychological) that causes the person disability or distress in social, personal, or occupational functioning.

Re: First physiological test for autism proves high accuracy in second trial

#49

Post by perl4ever »

icegreentea2 wrote: post #44
robrenaud wrote: post #37

Statistical reporting is often pretty frustrating. > “We are able to predict with 88 percent accuracy whether children have autism,” said Juergen Hahn, lead author > It is estimated that approximately 1.7 percent of all children are diagnosed with ASD def has_autism(child): return False apparently has 98% accuracy.

It's 88% correct classification on a validation data set consisting of 154 patients diagnosed with ASD. Their validation dataset did not contain a "typically developing" population since there's only one dataset for that data and it was consumed in their model generation. Their training data resulted in a 5% misclassification rate of their controls - that is 5% of patients that were not diagnosed with ASD in current…

If you give the test to everyone, your results are:

~4.9% incorrectly diagnosed with autism (0.05 x .983)

~1.5% correctly diagnosed with autism (0.88 x 0.017)

Thus, 76% of those diagnosed would be diagnosed incorrectly.

Post reply on HN